diff options
Diffstat (limited to 'Android.bp')
| -rw-r--r-- | Android.bp | 410 |
1 files changed, 205 insertions, 205 deletions
diff --git a/Android.bp b/Android.bp index c1adf32893..14ab3da246 100644 --- a/Android.bp +++ b/Android.bp @@ -50,7 +50,7 @@ cc_library_static { "src/core/NEON/kernels/assembly", "src/core/NEON/kernels/convolution/common", "src/core/NEON/kernels/convolution/winograd", - "src/core/cpu/kernels/assembly"], + "src/cpu/kernels/assembly"], export_include_dirs: [".", "./include"], srcs: [ "src/c/AclContext.cpp", @@ -245,145 +245,6 @@ cc_library_static { "src/core/Utils.cpp", "src/core/Validate.cpp", "src/core/Version.cpp", - "src/core/cpu/kernels/CpuActivationKernel.cpp", - "src/core/cpu/kernels/CpuAddKernel.cpp", - "src/core/cpu/kernels/CpuCastKernel.cpp", - "src/core/cpu/kernels/CpuCol2ImKernel.cpp", - "src/core/cpu/kernels/CpuConcatenateBatchKernel.cpp", - "src/core/cpu/kernels/CpuConcatenateDepthKernel.cpp", - "src/core/cpu/kernels/CpuConcatenateHeightKernel.cpp", - "src/core/cpu/kernels/CpuConcatenateWidthKernel.cpp", - "src/core/cpu/kernels/CpuConvertFullyConnectedWeightsKernel.cpp", - "src/core/cpu/kernels/CpuConvertQuantizedSignednessKernel.cpp", - "src/core/cpu/kernels/CpuCopyKernel.cpp", - "src/core/cpu/kernels/CpuDepthwiseConv2dNativeKernel.cpp", - "src/core/cpu/kernels/CpuDequantizeKernel.cpp", - "src/core/cpu/kernels/CpuDirectConv2dKernel.cpp", - "src/core/cpu/kernels/CpuDirectConv2dOutputStageKernel.cpp", - "src/core/cpu/kernels/CpuElementwiseKernel.cpp", - "src/core/cpu/kernels/CpuElementwiseUnaryKernel.cpp", - "src/core/cpu/kernels/CpuFillKernel.cpp", - "src/core/cpu/kernels/CpuFloorKernel.cpp", - "src/core/cpu/kernels/CpuGemmInterleave4x4Kernel.cpp", - "src/core/cpu/kernels/CpuGemmLowpMatrixMultiplyKernel.cpp", - "src/core/cpu/kernels/CpuGemmLowpMatrixReductionKernel.cpp", - "src/core/cpu/kernels/CpuGemmLowpOffsetContributionKernel.cpp", - "src/core/cpu/kernels/CpuGemmLowpOffsetContributionOutputStageKernel.cpp", - "src/core/cpu/kernels/CpuGemmLowpQuantizeDownInt32ScaleKernel.cpp", - "src/core/cpu/kernels/CpuGemmLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel.cpp", - "src/core/cpu/kernels/CpuGemmLowpQuantizeDownInt32ToInt8ScaleByFixedPointKernel.cpp", - "src/core/cpu/kernels/CpuGemmLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel.cpp", - "src/core/cpu/kernels/CpuGemmMatrixAdditionKernel.cpp", - "src/core/cpu/kernels/CpuGemmMatrixMultiplyKernel.cpp", - "src/core/cpu/kernels/CpuGemmTranspose1xWKernel.cpp", - "src/core/cpu/kernels/CpuIm2ColKernel.cpp", - "src/core/cpu/kernels/CpuMulKernel.cpp", - "src/core/cpu/kernels/CpuPermuteKernel.cpp", - "src/core/cpu/kernels/CpuPool2dKernel.cpp", - "src/core/cpu/kernels/CpuQuantizeKernel.cpp", - "src/core/cpu/kernels/CpuReshapeKernel.cpp", - "src/core/cpu/kernels/CpuScaleKernel.cpp", - "src/core/cpu/kernels/CpuSoftmaxKernel.cpp", - "src/core/cpu/kernels/CpuSubKernel.cpp", - "src/core/cpu/kernels/CpuTransposeKernel.cpp", - "src/core/cpu/kernels/CpuWeightsReshapeKernel.cpp", - "src/core/cpu/kernels/CpuWinogradConv2dKernel.cpp", - "src/core/cpu/kernels/activation/neon/fp16.cpp", - "src/core/cpu/kernels/activation/neon/fp32.cpp", - "src/core/cpu/kernels/activation/neon/qasymm8.cpp", - "src/core/cpu/kernels/activation/neon/qasymm8_signed.cpp", - "src/core/cpu/kernels/activation/neon/qsymm16.cpp", - "src/core/cpu/kernels/activation/sve/fp16.cpp", - "src/core/cpu/kernels/activation/sve/fp32.cpp", - "src/core/cpu/kernels/activation/sve/qasymm8.cpp", - "src/core/cpu/kernels/activation/sve/qasymm8_signed.cpp", - "src/core/cpu/kernels/activation/sve/qsymm16.cpp", - "src/core/cpu/kernels/add/neon/qasymm8.cpp", - "src/core/cpu/kernels/add/neon/qasymm8_signed.cpp", - "src/core/cpu/kernels/add/neon/qsymm16.cpp", - "src/core/cpu/kernels/add/sve/impl.cpp", - "src/core/cpu/kernels/add/sve/qasymm8.cpp", - "src/core/cpu/kernels/add/sve/qasymm8_signed.cpp", - "src/core/cpu/kernels/add/sve/qsymm16.cpp", - "src/core/cpu/kernels/elementwise/sve/elementwise.cpp", - "src/core/cpu/kernels/elementwise/sve/elementwise_unary.cpp", - "src/core/cpu/kernels/floor/neon/fp16.cpp", - "src/core/cpu/kernels/floor/neon/fp32.cpp", - "src/core/cpu/kernels/internal/CpuDepthwiseConv2dAssemblyWrapperKernel.cpp", - "src/core/cpu/kernels/internal/CpuPool2dAssemblyWrapperKernel.cpp", - "src/core/cpu/kernels/pool2d/neon/fp16.cpp", - "src/core/cpu/kernels/pool2d/neon/fp32.cpp", - "src/core/cpu/kernels/pool2d/neon/nchw/all.cpp", - "src/core/cpu/kernels/pool2d/neon/qasymm8.cpp", - "src/core/cpu/kernels/pool2d/neon/qasymm8_signed.cpp", - "src/core/cpu/kernels/scale/neon/fp16.cpp", - "src/core/cpu/kernels/scale/neon/integer.cpp", - "src/core/cpu/kernels/scale/neon/qasymm8.cpp", - "src/core/cpu/kernels/scale/neon/qasymm8_signed.cpp", - "src/core/cpu/kernels/scale/sve/fp16.cpp", - "src/core/cpu/kernels/scale/sve/fp32.cpp", - "src/core/cpu/kernels/scale/sve/integer.cpp", - "src/core/cpu/kernels/scale/sve/qasymm8.cpp", - "src/core/cpu/kernels/scale/sve/qasymm8_signed.cpp", - "src/core/cpu/kernels/softmax/impl/sve/impl.cpp", - "src/core/cpu/kernels/sub/neon/qasymm8.cpp", - "src/core/cpu/kernels/sub/neon/qasymm8_signed.cpp", - "src/core/cpu/kernels/sub/neon/qsymm16.cpp", - "src/core/gpu/cl/ClKernelLibrary.cpp", - "src/core/gpu/cl/kernels/ClActivationKernel.cpp", - "src/core/gpu/cl/kernels/ClBatchConcatenateKernel.cpp", - "src/core/gpu/cl/kernels/ClCastKernel.cpp", - "src/core/gpu/cl/kernels/ClCol2ImKernel.cpp", - "src/core/gpu/cl/kernels/ClConvertFullyConnectedWeightsKernel.cpp", - "src/core/gpu/cl/kernels/ClCopyKernel.cpp", - "src/core/gpu/cl/kernels/ClCropKernel.cpp", - "src/core/gpu/cl/kernels/ClDepthConcatenateKernel.cpp", - "src/core/gpu/cl/kernels/ClDequantizeKernel.cpp", - "src/core/gpu/cl/kernels/ClDirectConv2dKernel.cpp", - "src/core/gpu/cl/kernels/ClElementwiseKernel.cpp", - "src/core/gpu/cl/kernels/ClElementwiseUnaryKernel.cpp", - "src/core/gpu/cl/kernels/ClFillKernel.cpp", - "src/core/gpu/cl/kernels/ClFloorKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmLowpMatrixMultiplyNativeKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmLowpMatrixMultiplyReshapedKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmLowpOffsetContributionKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmLowpOffsetContributionOutputStageKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFixedPointKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmLowpReductionKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmMatrixMultiplyKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmMatrixMultiplyNativeKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmMatrixMultiplyReshapedKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmMatrixMultiplyReshapedOnlyRhsKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmReshapeLhsMatrixKernel.cpp", - "src/core/gpu/cl/kernels/ClGemmReshapeRhsMatrixKernel.cpp", - "src/core/gpu/cl/kernels/ClHeightConcatenateKernel.cpp", - "src/core/gpu/cl/kernels/ClIm2ColKernel.cpp", - "src/core/gpu/cl/kernels/ClMulKernel.cpp", - "src/core/gpu/cl/kernels/ClPermuteKernel.cpp", - "src/core/gpu/cl/kernels/ClPool2dKernel.cpp", - "src/core/gpu/cl/kernels/ClQuantizeKernel.cpp", - "src/core/gpu/cl/kernels/ClReshapeKernel.cpp", - "src/core/gpu/cl/kernels/ClScaleKernel.cpp", - "src/core/gpu/cl/kernels/ClSoftmaxKernel.cpp", - "src/core/gpu/cl/kernels/ClTransposeKernel.cpp", - "src/core/gpu/cl/kernels/ClWeightsReshapeKernel.cpp", - "src/core/gpu/cl/kernels/ClWidthConcatenate2TensorsKernel.cpp", - "src/core/gpu/cl/kernels/ClWidthConcatenate4TensorsKernel.cpp", - "src/core/gpu/cl/kernels/ClWidthConcatenateKernel.cpp", - "src/core/gpu/cl/kernels/ClWinogradFilterTransformKernel.cpp", - "src/core/gpu/cl/kernels/ClWinogradInputTransformKernel.cpp", - "src/core/gpu/cl/kernels/ClWinogradOutputTransformKernel.cpp", - "src/core/gpu/cl/kernels/gemm/ClGemmHelpers.cpp", - "src/core/gpu/cl/kernels/gemm/native/ClGemmDefaultConfigNativeBifrost.cpp", - "src/core/gpu/cl/kernels/gemm/native/ClGemmDefaultConfigNativeMidgard.cpp", - "src/core/gpu/cl/kernels/gemm/native/ClGemmDefaultConfigNativeValhall.cpp", - "src/core/gpu/cl/kernels/gemm/reshaped/ClGemmDefaultConfigReshapedBifrost.cpp", - "src/core/gpu/cl/kernels/gemm/reshaped/ClGemmDefaultConfigReshapedValhall.cpp", - "src/core/gpu/cl/kernels/gemm/reshaped_only_rhs/ClGemmDefaultConfigReshapedRhsOnlyBifrost.cpp", - "src/core/gpu/cl/kernels/gemm/reshaped_only_rhs/ClGemmDefaultConfigReshapedRhsOnlyValhall.cpp", "src/core/helpers/SoftmaxHelpers.cpp", "src/core/helpers/WindowHelpers.cpp", "src/core/utils/AssemblyUtils.cpp", @@ -400,9 +261,213 @@ cc_library_static { "src/cpu/CpuContext.cpp", "src/cpu/CpuQueue.cpp", "src/cpu/CpuTensor.cpp", + "src/cpu/kernels/CpuActivationKernel.cpp", + "src/cpu/kernels/CpuAddKernel.cpp", + "src/cpu/kernels/CpuCastKernel.cpp", + "src/cpu/kernels/CpuCol2ImKernel.cpp", + "src/cpu/kernels/CpuConcatenateBatchKernel.cpp", + "src/cpu/kernels/CpuConcatenateDepthKernel.cpp", + "src/cpu/kernels/CpuConcatenateHeightKernel.cpp", + "src/cpu/kernels/CpuConcatenateWidthKernel.cpp", + "src/cpu/kernels/CpuConvertFullyConnectedWeightsKernel.cpp", + "src/cpu/kernels/CpuConvertQuantizedSignednessKernel.cpp", + "src/cpu/kernels/CpuCopyKernel.cpp", + "src/cpu/kernels/CpuDepthwiseConv2dNativeKernel.cpp", + "src/cpu/kernels/CpuDequantizeKernel.cpp", + "src/cpu/kernels/CpuDirectConv2dKernel.cpp", + "src/cpu/kernels/CpuDirectConv2dOutputStageKernel.cpp", + "src/cpu/kernels/CpuElementwiseKernel.cpp", + "src/cpu/kernels/CpuElementwiseUnaryKernel.cpp", + "src/cpu/kernels/CpuFillKernel.cpp", + "src/cpu/kernels/CpuFloorKernel.cpp", + "src/cpu/kernels/CpuGemmInterleave4x4Kernel.cpp", + "src/cpu/kernels/CpuGemmLowpMatrixMultiplyKernel.cpp", + "src/cpu/kernels/CpuGemmLowpMatrixReductionKernel.cpp", + "src/cpu/kernels/CpuGemmLowpOffsetContributionKernel.cpp", + "src/cpu/kernels/CpuGemmLowpOffsetContributionOutputStageKernel.cpp", + "src/cpu/kernels/CpuGemmLowpQuantizeDownInt32ScaleKernel.cpp", + "src/cpu/kernels/CpuGemmLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel.cpp", + "src/cpu/kernels/CpuGemmLowpQuantizeDownInt32ToInt8ScaleByFixedPointKernel.cpp", + "src/cpu/kernels/CpuGemmLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel.cpp", + "src/cpu/kernels/CpuGemmMatrixAdditionKernel.cpp", + "src/cpu/kernels/CpuGemmMatrixMultiplyKernel.cpp", + "src/cpu/kernels/CpuGemmTranspose1xWKernel.cpp", + "src/cpu/kernels/CpuIm2ColKernel.cpp", + "src/cpu/kernels/CpuMulKernel.cpp", + "src/cpu/kernels/CpuPermuteKernel.cpp", + "src/cpu/kernels/CpuPool2dKernel.cpp", + "src/cpu/kernels/CpuQuantizeKernel.cpp", + "src/cpu/kernels/CpuReshapeKernel.cpp", + "src/cpu/kernels/CpuScaleKernel.cpp", + "src/cpu/kernels/CpuSoftmaxKernel.cpp", + "src/cpu/kernels/CpuSubKernel.cpp", + "src/cpu/kernels/CpuTransposeKernel.cpp", + "src/cpu/kernels/CpuWeightsReshapeKernel.cpp", + "src/cpu/kernels/CpuWinogradConv2dKernel.cpp", + "src/cpu/kernels/activation/neon/fp16.cpp", + "src/cpu/kernels/activation/neon/fp32.cpp", + "src/cpu/kernels/activation/neon/qasymm8.cpp", + "src/cpu/kernels/activation/neon/qasymm8_signed.cpp", + "src/cpu/kernels/activation/neon/qsymm16.cpp", + "src/cpu/kernels/activation/sve/fp16.cpp", + "src/cpu/kernels/activation/sve/fp32.cpp", + "src/cpu/kernels/activation/sve/qasymm8.cpp", + "src/cpu/kernels/activation/sve/qasymm8_signed.cpp", + "src/cpu/kernels/activation/sve/qsymm16.cpp", + "src/cpu/kernels/add/neon/qasymm8.cpp", + "src/cpu/kernels/add/neon/qasymm8_signed.cpp", + "src/cpu/kernels/add/neon/qsymm16.cpp", + "src/cpu/kernels/add/sve/impl.cpp", + "src/cpu/kernels/add/sve/qasymm8.cpp", + "src/cpu/kernels/add/sve/qasymm8_signed.cpp", + "src/cpu/kernels/add/sve/qsymm16.cpp", + "src/cpu/kernels/elementwise/sve/elementwise.cpp", + "src/cpu/kernels/elementwise/sve/elementwise_unary.cpp", + "src/cpu/kernels/floor/neon/fp16.cpp", + "src/cpu/kernels/floor/neon/fp32.cpp", + "src/cpu/kernels/internal/CpuDepthwiseConv2dAssemblyWrapperKernel.cpp", + "src/cpu/kernels/internal/CpuPool2dAssemblyWrapperKernel.cpp", + "src/cpu/kernels/pool2d/neon/fp16.cpp", + "src/cpu/kernels/pool2d/neon/fp32.cpp", + "src/cpu/kernels/pool2d/neon/nchw/all.cpp", + "src/cpu/kernels/pool2d/neon/qasymm8.cpp", + "src/cpu/kernels/pool2d/neon/qasymm8_signed.cpp", + "src/cpu/kernels/scale/neon/fp16.cpp", + "src/cpu/kernels/scale/neon/integer.cpp", + "src/cpu/kernels/scale/neon/qasymm8.cpp", + "src/cpu/kernels/scale/neon/qasymm8_signed.cpp", + "src/cpu/kernels/scale/sve/fp16.cpp", + "src/cpu/kernels/scale/sve/fp32.cpp", + "src/cpu/kernels/scale/sve/integer.cpp", + "src/cpu/kernels/scale/sve/qasymm8.cpp", + "src/cpu/kernels/scale/sve/qasymm8_signed.cpp", + "src/cpu/kernels/softmax/impl/sve/impl.cpp", + "src/cpu/kernels/sub/neon/qasymm8.cpp", + "src/cpu/kernels/sub/neon/qasymm8_signed.cpp", + "src/cpu/kernels/sub/neon/qsymm16.cpp", + "src/cpu/operators/CpuActivation.cpp", + "src/cpu/operators/CpuAdd.cpp", + "src/cpu/operators/CpuCast.cpp", + "src/cpu/operators/CpuConcatenate.cpp", + "src/cpu/operators/CpuConv2d.cpp", + "src/cpu/operators/CpuConvertFullyConnectedWeights.cpp", + "src/cpu/operators/CpuCopy.cpp", + "src/cpu/operators/CpuDepthwiseConv2d.cpp", + "src/cpu/operators/CpuDepthwiseConv2dAssemblyDispatch.cpp", + "src/cpu/operators/CpuDequantize.cpp", + "src/cpu/operators/CpuDirectConv2d.cpp", + "src/cpu/operators/CpuElementwise.cpp", + "src/cpu/operators/CpuElementwiseUnary.cpp", + "src/cpu/operators/CpuFill.cpp", + "src/cpu/operators/CpuFlatten.cpp", + "src/cpu/operators/CpuFloor.cpp", + "src/cpu/operators/CpuFullyConnected.cpp", + "src/cpu/operators/CpuGemm.cpp", + "src/cpu/operators/CpuGemmConv2d.cpp", + "src/cpu/operators/CpuGemmDirectConv2d.cpp", + "src/cpu/operators/CpuGemmLowpMatrixMultiplyCore.cpp", + "src/cpu/operators/CpuGemmLowpOutputStage.cpp", + "src/cpu/operators/CpuMul.cpp", + "src/cpu/operators/CpuPermute.cpp", + "src/cpu/operators/CpuPool2d.cpp", + "src/cpu/operators/CpuQuantize.cpp", + "src/cpu/operators/CpuReshape.cpp", + "src/cpu/operators/CpuScale.cpp", + "src/cpu/operators/CpuSoftmax.cpp", + "src/cpu/operators/CpuSub.cpp", + "src/cpu/operators/CpuTranspose.cpp", + "src/cpu/operators/CpuWinogradConv2d.cpp", + "src/cpu/operators/internal/CpuGemmAssemblyDispatch.cpp", "src/gpu/cl/ClContext.cpp", + "src/gpu/cl/ClKernelLibrary.cpp", "src/gpu/cl/ClQueue.cpp", "src/gpu/cl/ClTensor.cpp", + "src/gpu/cl/kernels/ClActivationKernel.cpp", + "src/gpu/cl/kernels/ClBatchConcatenateKernel.cpp", + "src/gpu/cl/kernels/ClCastKernel.cpp", + "src/gpu/cl/kernels/ClCol2ImKernel.cpp", + "src/gpu/cl/kernels/ClConvertFullyConnectedWeightsKernel.cpp", + "src/gpu/cl/kernels/ClCopyKernel.cpp", + "src/gpu/cl/kernels/ClCropKernel.cpp", + "src/gpu/cl/kernels/ClDepthConcatenateKernel.cpp", + "src/gpu/cl/kernels/ClDequantizeKernel.cpp", + "src/gpu/cl/kernels/ClDirectConv2dKernel.cpp", + "src/gpu/cl/kernels/ClElementwiseKernel.cpp", + "src/gpu/cl/kernels/ClElementwiseUnaryKernel.cpp", + "src/gpu/cl/kernels/ClFillKernel.cpp", + "src/gpu/cl/kernels/ClFloorKernel.cpp", + "src/gpu/cl/kernels/ClGemmLowpMatrixMultiplyNativeKernel.cpp", + "src/gpu/cl/kernels/ClGemmLowpMatrixMultiplyReshapedKernel.cpp", + "src/gpu/cl/kernels/ClGemmLowpMatrixMultiplyReshapedOnlyRhsKernel.cpp", + "src/gpu/cl/kernels/ClGemmLowpOffsetContributionKernel.cpp", + "src/gpu/cl/kernels/ClGemmLowpOffsetContributionOutputStageKernel.cpp", + "src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFixedPointKernel.cpp", + "src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleByFloatKernel.cpp", + "src/gpu/cl/kernels/ClGemmLowpQuantizeDownInt32ScaleKernel.cpp", + "src/gpu/cl/kernels/ClGemmLowpReductionKernel.cpp", + "src/gpu/cl/kernels/ClGemmMatrixMultiplyKernel.cpp", + "src/gpu/cl/kernels/ClGemmMatrixMultiplyNativeKernel.cpp", + "src/gpu/cl/kernels/ClGemmMatrixMultiplyReshapedKernel.cpp", + "src/gpu/cl/kernels/ClGemmMatrixMultiplyReshapedOnlyRhsKernel.cpp", + "src/gpu/cl/kernels/ClGemmReshapeLhsMatrixKernel.cpp", + "src/gpu/cl/kernels/ClGemmReshapeRhsMatrixKernel.cpp", + "src/gpu/cl/kernels/ClHeightConcatenateKernel.cpp", + "src/gpu/cl/kernels/ClIm2ColKernel.cpp", + "src/gpu/cl/kernels/ClMulKernel.cpp", + "src/gpu/cl/kernels/ClPermuteKernel.cpp", + "src/gpu/cl/kernels/ClPool2dKernel.cpp", + "src/gpu/cl/kernels/ClQuantizeKernel.cpp", + "src/gpu/cl/kernels/ClReshapeKernel.cpp", + "src/gpu/cl/kernels/ClScaleKernel.cpp", + "src/gpu/cl/kernels/ClSoftmaxKernel.cpp", + "src/gpu/cl/kernels/ClTransposeKernel.cpp", + "src/gpu/cl/kernels/ClWeightsReshapeKernel.cpp", + "src/gpu/cl/kernels/ClWidthConcatenate2TensorsKernel.cpp", + "src/gpu/cl/kernels/ClWidthConcatenate4TensorsKernel.cpp", + "src/gpu/cl/kernels/ClWidthConcatenateKernel.cpp", + "src/gpu/cl/kernels/ClWinogradFilterTransformKernel.cpp", + "src/gpu/cl/kernels/ClWinogradInputTransformKernel.cpp", + "src/gpu/cl/kernels/ClWinogradOutputTransformKernel.cpp", + "src/gpu/cl/kernels/gemm/ClGemmHelpers.cpp", + "src/gpu/cl/kernels/gemm/native/ClGemmDefaultConfigNativeBifrost.cpp", + "src/gpu/cl/kernels/gemm/native/ClGemmDefaultConfigNativeMidgard.cpp", + "src/gpu/cl/kernels/gemm/native/ClGemmDefaultConfigNativeValhall.cpp", + "src/gpu/cl/kernels/gemm/reshaped/ClGemmDefaultConfigReshapedBifrost.cpp", + "src/gpu/cl/kernels/gemm/reshaped/ClGemmDefaultConfigReshapedValhall.cpp", + "src/gpu/cl/kernels/gemm/reshaped_only_rhs/ClGemmDefaultConfigReshapedRhsOnlyBifrost.cpp", + "src/gpu/cl/kernels/gemm/reshaped_only_rhs/ClGemmDefaultConfigReshapedRhsOnlyValhall.cpp", + "src/gpu/cl/operators/ClActivation.cpp", + "src/gpu/cl/operators/ClAdd.cpp", + "src/gpu/cl/operators/ClCast.cpp", + "src/gpu/cl/operators/ClConcatenate.cpp", + "src/gpu/cl/operators/ClConv2d.cpp", + "src/gpu/cl/operators/ClConvertFullyConnectedWeights.cpp", + "src/gpu/cl/operators/ClCopy.cpp", + "src/gpu/cl/operators/ClCrop.cpp", + "src/gpu/cl/operators/ClDequantize.cpp", + "src/gpu/cl/operators/ClDirectConv2d.cpp", + "src/gpu/cl/operators/ClElementwiseOperations.cpp", + "src/gpu/cl/operators/ClElementwiseUnary.cpp", + "src/gpu/cl/operators/ClFill.cpp", + "src/gpu/cl/operators/ClFlatten.cpp", + "src/gpu/cl/operators/ClFloor.cpp", + "src/gpu/cl/operators/ClFullyConnected.cpp", + "src/gpu/cl/operators/ClGemm.cpp", + "src/gpu/cl/operators/ClGemmConv2d.cpp", + "src/gpu/cl/operators/ClGemmLowpMatrixMultiplyCore.cpp", + "src/gpu/cl/operators/ClGemmLowpOutputStage.cpp", + "src/gpu/cl/operators/ClLogicalNot.cpp", + "src/gpu/cl/operators/ClMul.cpp", + "src/gpu/cl/operators/ClPRelu.cpp", + "src/gpu/cl/operators/ClPermute.cpp", + "src/gpu/cl/operators/ClPool2d.cpp", + "src/gpu/cl/operators/ClQuantize.cpp", + "src/gpu/cl/operators/ClReshape.cpp", + "src/gpu/cl/operators/ClScale.cpp", + "src/gpu/cl/operators/ClSoftmax.cpp", + "src/gpu/cl/operators/ClSub.cpp", + "src/gpu/cl/operators/ClTranspose.cpp", + "src/gpu/cl/operators/ClWinogradConv2d.cpp", "src/runtime/Allocator.cpp", "src/runtime/BlobLifetimeManager.cpp", "src/runtime/BlobMemoryPool.cpp", @@ -629,71 +694,6 @@ cc_library_static { "src/runtime/Tensor.cpp", "src/runtime/TensorAllocator.cpp", "src/runtime/Utils.cpp", - "src/runtime/cpu/operators/CpuActivation.cpp", - "src/runtime/cpu/operators/CpuAdd.cpp", - "src/runtime/cpu/operators/CpuCast.cpp", - "src/runtime/cpu/operators/CpuConcatenate.cpp", - "src/runtime/cpu/operators/CpuConv2d.cpp", - "src/runtime/cpu/operators/CpuConvertFullyConnectedWeights.cpp", - "src/runtime/cpu/operators/CpuCopy.cpp", - "src/runtime/cpu/operators/CpuDepthwiseConv2d.cpp", - "src/runtime/cpu/operators/CpuDepthwiseConv2dAssemblyDispatch.cpp", - "src/runtime/cpu/operators/CpuDequantize.cpp", - "src/runtime/cpu/operators/CpuDirectConv2d.cpp", - "src/runtime/cpu/operators/CpuElementwise.cpp", - "src/runtime/cpu/operators/CpuElementwiseUnary.cpp", - "src/runtime/cpu/operators/CpuFill.cpp", - "src/runtime/cpu/operators/CpuFlatten.cpp", - "src/runtime/cpu/operators/CpuFloor.cpp", - "src/runtime/cpu/operators/CpuFullyConnected.cpp", - "src/runtime/cpu/operators/CpuGemm.cpp", - "src/runtime/cpu/operators/CpuGemmConv2d.cpp", - "src/runtime/cpu/operators/CpuGemmDirectConv2d.cpp", - "src/runtime/cpu/operators/CpuGemmLowpMatrixMultiplyCore.cpp", - "src/runtime/cpu/operators/CpuGemmLowpOutputStage.cpp", - "src/runtime/cpu/operators/CpuMul.cpp", - "src/runtime/cpu/operators/CpuPermute.cpp", - "src/runtime/cpu/operators/CpuPool2d.cpp", - "src/runtime/cpu/operators/CpuQuantize.cpp", - "src/runtime/cpu/operators/CpuReshape.cpp", - "src/runtime/cpu/operators/CpuScale.cpp", - "src/runtime/cpu/operators/CpuSoftmax.cpp", - "src/runtime/cpu/operators/CpuSub.cpp", - "src/runtime/cpu/operators/CpuTranspose.cpp", - "src/runtime/cpu/operators/CpuWinogradConv2d.cpp", - "src/runtime/cpu/operators/internal/CpuGemmAssemblyDispatch.cpp", - "src/runtime/gpu/cl/operators/ClActivation.cpp", - "src/runtime/gpu/cl/operators/ClAdd.cpp", - "src/runtime/gpu/cl/operators/ClCast.cpp", - "src/runtime/gpu/cl/operators/ClConcatenate.cpp", - "src/runtime/gpu/cl/operators/ClConv2d.cpp", - "src/runtime/gpu/cl/operators/ClConvertFullyConnectedWeights.cpp", - "src/runtime/gpu/cl/operators/ClCopy.cpp", - "src/runtime/gpu/cl/operators/ClCrop.cpp", - "src/runtime/gpu/cl/operators/ClDequantize.cpp", - "src/runtime/gpu/cl/operators/ClDirectConv2d.cpp", - "src/runtime/gpu/cl/operators/ClElementwiseOperations.cpp", - "src/runtime/gpu/cl/operators/ClElementwiseUnary.cpp", - "src/runtime/gpu/cl/operators/ClFill.cpp", - "src/runtime/gpu/cl/operators/ClFlatten.cpp", - "src/runtime/gpu/cl/operators/ClFloor.cpp", - "src/runtime/gpu/cl/operators/ClFullyConnected.cpp", - "src/runtime/gpu/cl/operators/ClGemm.cpp", - "src/runtime/gpu/cl/operators/ClGemmConv2d.cpp", - "src/runtime/gpu/cl/operators/ClGemmLowpMatrixMultiplyCore.cpp", - "src/runtime/gpu/cl/operators/ClGemmLowpOutputStage.cpp", - "src/runtime/gpu/cl/operators/ClLogicalNot.cpp", - "src/runtime/gpu/cl/operators/ClMul.cpp", - "src/runtime/gpu/cl/operators/ClPRelu.cpp", - "src/runtime/gpu/cl/operators/ClPermute.cpp", - "src/runtime/gpu/cl/operators/ClPool2d.cpp", - "src/runtime/gpu/cl/operators/ClQuantize.cpp", - "src/runtime/gpu/cl/operators/ClReshape.cpp", - "src/runtime/gpu/cl/operators/ClScale.cpp", - "src/runtime/gpu/cl/operators/ClSoftmax.cpp", - "src/runtime/gpu/cl/operators/ClSub.cpp", - "src/runtime/gpu/cl/operators/ClTranspose.cpp", - "src/runtime/gpu/cl/operators/ClWinogradConv2d.cpp", "utils/CommonGraphOptions.cpp", "utils/GraphUtils.cpp", "utils/Utils.cpp", |